PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
October 2014
929
Q1: Can I identify stream channel boundaries (rivers and
creeks) accurately from 0.8 average point spacing?
Dr. Abdullah:
The short answer to your question is yes; however, accuracy here
is subjective as it depends on which accuracy that you are trying
to achieve. If your term of “accurately” refers to the quality of
identification, then yes, LiDAR with this kind of density should
give you fairly good details of the water boundaries for most
of the water channels and streams unless they are too small
to be spotted by LiDAR. Figure 1 is a good representation of
the quality of LiDAR in defining water boundaries. The LiDAR
image in Figure 1 clearly shows the fine details of a drainage
network that can easily be used for delineating hydro features.
However, if you re-
fer to geometrical
accuracy of delinea-
tion, then the accu-
racy will be limited
to the horizontal
accuracy of your Li-
DAR dataset. Hor-
izontal accuracy of
LiDAR dataset is
influenced by the
technical specifica-
tions of the LiDAR
system you use for
acquiring the data-
set. Different Li-
DAR systems have different aperture size and different flying
altitude and therefore different pulse divergence on the ground.
In many cases, pulse divergence (or the illuminated footprint di-
ameter) is used simplistically to estimate horizontal accuracy. A
practical estimate for the horizontal accuracy of LiDAR dataset
based on the size of the illuminated footprint is:
Horizontal Accuracy (
1
σ
) = ½ the illuminated footprint diameter
A more accurate estimate of LiDAR horizontal accuracy is
achieved by error propagation or modeling the accuracy of the
inertial measurement unit (IMU) used for the system attitude
determination, the on-board GPS used for system positioning
in space and other system errors. All commercial LiDAR sys-
tems available in the market today provide accurate horizon-
The following are short answers for questions raised by participants of
the ASPRS online webinar “LiDAR Fundamental and Applications”.
tal accuracy estimates during the mission planning stage. A
highly accurate LiDAR dataset with a post spacing of 1 m is
expected to have a horizontal accuracy of around 0.20 m (1σ).
Q2: Which return depicts tree height (what happens if a
return hits the side of tree and not absolute top)?
Dr. Abdullah
The first return of the LiDAR pulse is usually associated with
highest elevated natural features (such as trees) or man-made
objects (such as buildings) on the ground. A pulse’s return
maps the position where it is reflected from; therefore, if the
first return is reflected from the side of the tree and not the
top, then you will be looking at the side of tree and not the
top. Here I would like to clarify that the unique quality about
the LiDAR dataset set is its high density. Having a few pulses
return from the side of a tree does not mean that you will
not be able to map the top of the foliage (unless you are after
an individual tree). When viewing the first round of returns
collectively, the group clearly defines the top of the foliage line
despite the fact that some of the returns are reflected from
lower branches (Figure 2). From Figure 2, you notice a clear
signature of the tree tops despite the fact that some LiDAR
returns are reflected from within the foliage.
**Dr. Abdullah is Senior Geospatial Scientist at Woolpert,
Inc. He is the 2010 recipient of the ASPRS Photogrammetric
(Fairchild) Award.
The contents of this column reflect the views of the author,
who is responsible for the facts and accuracy of the data pre-
sented herein. The contents do not necessarily reflect the offi-
cial views or policies of the American Society for Photogram-
metry and Remote Sensing and/or Woolpert, Inc.
Figure 1 Water boundaries in LiDAR
dataset
Figure 2 Top of trees line from LiDAR dataset
